Public Protocols & Systems
Deep Bound Research is committed to open-source infrastructure for regulated AI autonomy.
Release Readiness
Selected public repositories and technical artifacts will be published as they pass documentation, safety, and release-readiness review.
Public release follows a staged review process designed to balance transparency, reproducibility, and operational safety.
02 / Repository Index
ACE Context Engine
A retrieval and context-ranking engine for task-aware AI execution.
- Visibility
- Open Source Preparation
- Release
- Open source preparation
- License
- MIT
- System
- ACE
Boundary Evaluation Harness
Benchmark suites for evaluating agent governance protocols.
- Visibility
- Planned
- Release
- Planned
- License
- Apache-2.0
- System
- Boundary
Regulated-Actions Protocol
A protocol specification for safe tool-usage by autonomous agents.
- Visibility
- Public Preview
- Release
- Public link pending
- License
- CC-BY-4.0
- System
- Cerberus
Documentation
Related Public Artifacts
Governance protocols, technical notes, and system overviews released for public review. Full index available on the artifacts page.
Eve Constitution
Runtime Governance Constitution · Ex1
ACE Open Source Plan
Open Source Plan · ACE
Ex1 Operator Runtime Overview
System Overview · Ex1
Boundary Trace Model
Technical Note · Boundary
Cerberus Defensive Harness Note
Technical Note · Cerberus
StrongHold Ingest & Archive Technical Note
Technical Note · StrongHold
Release Policy
We publish public-safe governance protocols, evaluation harnesses, and technical notes before exposing implementation details. Systems that remain in private research phases are summarized at the capability and principle level until they are stable, reviewed, and safe to release.
Commitment to Open Governance
While many of our core systems remain in private research phases to ensure safety and stability, we aim to release the governance protocols and evaluation harnesses that underpin them as open standards.